基于图像处理的课室人员位置分布识别

对大型课室人员位置分布识别有利于课室灯光节能管理,基于面积判据的人员位置识别算法没有应用场景的约束,但容易因人体Blob图像粘连而产生误判;基于座位网格的算法只适用于直线排列的固定座位,但算法简单可靠。实验证明:两种方法均能有效地检测出人员分布情况,后者对固定座位的课室检测准确性更好,效率更高。

People Distribution Recognizing in Classroom Based on Image Processing

People Distribution recognizing in large classroom is used for power saving and lighting management. The people distribution recognizing algorithm based on area criterion is short of constraints in application scenes, but misjudgment will be made with this algorithm when the people’s blob image are too closed. Another algorithm based on classroom seats grid criterion only applies to fixed linear arrangement seats; the algorithm is simple and reliable. Experiments prove that the two algorithms can recognize the distribution of people availably both. And the algorithm based on seats grid criterion is more accurate and efficient in the situation of fixed linear arrangement seats classroom.